Anomalous functional language lateralization in semantic variant PPA

In 1982, Marsel Mesulam detailed a series of cases with primary progressive aphasia (PPA). Originally defined, PPA was restricted to patients with focal left hemisphere involvement. As cases with right hemisphere disease appeared,^1–3 an amended theory proposed that PPA arose from a “locus of least resistance” within the dominant language network.⁴ Ultimately, this explanation was unable to account for a specific subtype of semantic variant PPA (svPPA) where pathology first involved the nondominant hemisphere and subsequently attacked the dominant anterior temporal lobe language network.⁵ With the discovery of increased non-right-handedness (nRH) and planum temporale symmetry within our svPPA cohort,⁶ we put forward a refinement of the prevailing hypothesis and suggested that mechanisms of lateralization and neurodevelopment that shape the language network also generate a differential pattern of disease susceptibility within specific PPA variants. We postulated that the increased nRH observed in svPPA might be secondary to an increased incidence of anomalous language laterality. In this report, we describe the magnetoencephalography (MEG) pattern of language lateralization in a single svPPA case with left temporal predominant atrophy despite a clinical presentation of a right temporal syndrome.

Methods.

We obtained written patient consent and study approval by the UCSF Institutional Review Board. We performed comprehensive neurologic, neuropsychological, language battery,⁶ brain MRI, FDG and Pittsburgh compound B (PiB) PET imaging, and genetic testing.

MEG data were collected using a 275-channel whole-head biomagnetometer (CTF; Coquitlam, BC, Canada) during an auditory verb generation task.⁷ MEG sensor data were reconstructed in brain source space using a time-frequency optimized adaptive spatial filtering technique (nutmeg.berkeley.edu). Oscillatory power changes in the beta (12–30 Hz) and high-gamma (50–120 Hz) bands were estimated using a pseudo-F statistic, comparing sliding active windows to an intertrail baseline period. MEG imaging reconstructions were analyzed time locked either to stimulus onset (0 ms = auditory noun presentation) or to response onset (0 ms = vocal verb generation), and a laterality index was derived as a ratio of beta power change between the left and right hemispheres over temporal and frontal regions.⁷

Case report.

We present data from a 61-year-old ambidextrous man with 8 years of progressive behavioral changes evidenced by new onset of hyper-religious dogmatism and inability to recognize familiar faces. He developed a new fascination with words, manifested as hyperverbalism and obsession with word games, and rigidly sought out specific foods. A year prior to presentation, he began substituting words, “that boy” for “son,” “temper” for “temperature,” and “knob” for “ankle.” On examination, he revealed pressured, fluent speech, and spoke in rhymes and puns. Edinburgh Handedness Inventory corroborated ambidexterity (+0.3). Neuropsychological testing showed intact global cognition (30/30 Folstein Mini-Mental State Examination), verbal memory (average for age on California Verbal Learning Task short form), and visuospatial abilities (15/17 Benson Figure copy). He performed poorly on Stroop interference executive testing and displayed greatest impairments on confrontation naming (8/15 Boston Naming Task) and face recognition (0/12 Famous Faces).

Results.

Genetics.

C9ORF72, GRN, and TDP-43 mutations were absent.

PET/MRI.

FDG-PET showed relatively symmetrical hypometabolism in the temporal poles, while PiB-PET was negative for cortical amyloid. T1-weighted MRI revealed prominent, left greater than right, anterior temporal lobe atrophy extending into the insulae without appreciable frontal involvement.

MEG.

Robust beta-band power suppression was observed over the right hemisphere in the svPPA case in both stimulus-locked and response-locked verb generation tasks, in contrast to the highly left-lateralized activation observed in healthy individuals⁷ (figure). Quantitative comparison of activity across the left and right hemispheres produced a laterality index of −0.46, indicative of right hemisphere language dominance.

Figure. Stimulus-locked and response-locked magnetoencephalography activation patterns.

(A) Changes in β band (12–30 Hz) oscillatory activity during the speech reception phase (verb generation task, stimulus-locked) for an average of 18 healthy controls (left column) and the semantic variant primary progressive aphasia (svPPA) case (right column). Increases in beta power suppression during later time windows (550–750 ms, poststimulus) are largely left hemisphere dominant in the control cohort. Increased right hemisphere beta suppression is observed in the svPPA case for the same time windows. (B) Changes in beta band (12–30 Hz) oscillatory activity during the speech production phase (verb generation task, response-locked) for an average of 18 healthy controls (left column) and the svPPA case (right column). Increases in beta power suppression prior to response onset (−550 ms) are largely left hemisphere dominant in the control cohort and right hemisphere dominant in the svPPA case. MEG = magnetoencephalography.

Discussion.

We report the functional identification of right hemisphere language dominance in an atypical right temporal svPPA clinical presentation with predominant left-sided anterior temporal lobe atrophy. Previously, anomalous language lateralization was reported in a left-handed PPA patient with right-sided, not left-sided, atrophy.³ In both cases, discordant clinical and radiologic observations prompted functional imaging assessments, and the presence of anomalous language lateralization reconciled these observations. Whether anomalous language activity is secondary to neurodevelopment or is in response to the disease process is uncertain and needs to be addressed in future studies. Nevertheless, the clinical presentation and nRH status of both supports a history of premorbid right hemispheric language dominance. Anomalous language lateralization is suspected in PPA when hypometabolism or atrophy pattern conflicts with clinical predictions.^1–3 Larger cohort studies may be able to adequately address the issue of whether or not anomalous lateralization is more common in PPA, or specific subtypes of PPA, than the general population. Approaches like MEG (potentially when combined with other noninvasive imaging measures) may be important and effective research tools for investigating the natural history of language lateralization in PPA.